Background: Previous studies applying Sepsis-3 criteria to pediatric sepsis were based on retrospective analyses of pediatric intensive care unit (PICU) cohorts not designed to investigate sepsis and included children without microbiologically defined infections. We aimed to validate organ dysfunction criteria in a prospective cohort of children with blood culture-proven sepsis, including emergency department, PICU, and ward patients.
Methods: Between 1.9.2011 and 31.12.2015, this multi-center prospective population-based cohort study recruited children < 17 years with blood culture-proven sepsis defined according to 2005 International Pediatric Sepsis Consensus Conference (IPSCC) criteria. We excluded prematurely born infants and term born infants aged ≤ 7 days. We compared the performance of 2005 IPSCC, Pediatric Logistic Organ Dysfunction (PELOD)-2, and pediatric Sequential Organ Failure Assessment (pSOFA) score on the day of blood culture sampling to identify children who died ≤ 30 days after sepsis onset using ROC curves. To adjust 2005 IPSCC, PELOD-2, and pSOFA for age, sex, and presence of chronic medical conditions, we fitted logistic mixed-models.
Results: We analyzed 878 sepsis episodes in 807 children. In 288 (33%) episodes an organ dysfunction was present. Thirty-eight (4%) children died within 30 days after sepsis onset. Case fatality ratio was 1% in children without organ dysfunction, 5% in children with a single organ dysfunction, and 21% in children with multi organ dysfunction. The AUC to discriminate sepsis episodes with fatal outcome was 0.82 (95% CI 0.74 - 0.90) for 2005 IPSCC, 0.73 (0.63 - 0.83) for PELOD-2, and 0.78 (0.69 - 0.88) for pSOFA, with adjusted AUCs of 0.87 (0.82 - 0.92), 0.83 (0.76 - 0.89), and 0.85 (0.78 - 0.92). Cardiac, respiratory, and neurological dysfunction were the most relevant organ dysfunctions in our population. Considering only these three organs AUCs were 0.78 (0.70 - 0.87) for 2005 IPSCC, 0.72 (0.62 - 0.82) for PELOD-2, and 0.76 (0.66 - 0.86) for pSOFA, with adjusted AUCs of 0.9 (0.85 - 0.94) for 2005 IPSCC, 0.87 (0.82 - 0.93) for PELOD-2, and 0.88 (0.82 - 0.94) for pSFOA.
Conclusions: Different measures of pediatric organ dysfunction performed comparably, which can inform the revision of pediatric sepsis definitions. While presence of any organ failure increased the risk of death in pediatric sepsis, cardiac, pulmonary, and neurological dysfunction were most important with respect to death.